IP Library Granted Patent US 12,632,530
Granted Patent B2
US 12,632,530 · App. 18/856,678 · Granted May 19, 2026

Method for securing a biometric recognition of a user

Inventor: Jan Vacek (Prague, CZ)
Assignee: THALES DIS FRANCE SAS
G06F21/32
View Patent ↗
Loading inventors, assignments & file history…
Monitor This Case
Get email alerts when status or documents change.
Order Certified Copies
Most orders are placed with the USPTO same day — all within 24 business hours.
Order via The Patent Place →
Pre-filled with this patent's details
Quick Facts
Patent No.
US 12,632,530
App. No.
18/856,678
Granted
May 19, 2026
Kind
B2
Abstract

Provided is a method for securing a biometric recognition of a user to be recognized against one or more templates and a non zero integer, corresponding to registered users identities, using a biometric sample of said user, in a biometric recognition system comprising a plurality of computation devices, a global scoring computation device and a recognition device. Each of said templates is split, using an additive splitting modulo a predetermined integer N, into a plurality of template shares stored in said computation devices, with i in [0, n−1] and n a non zero integer. A global scoring is computed for said at least one template by combining said shares of the global scoring computed for said template by said plurality of computation devices according to a formula. Other embodiments disclosed.

Claims (173)

1 . A method comprising steps of:

securing a biometric recognition of a user to be recognized against one or more templates T j , with j in [0, J−1] and J a non zero integer, corresponding to registered users identities;

using a biometric sample of said user, in a biometric recognition system comprising a plurality of computation devices, a global scoring computation device and a recognition device;

splitting each of said templates using an additive splitting modulo a predetermined integer N, into a plurality of template shares

T

i

j

stored in said computation devices, with i in [0, n−1] and n a non zero integer, wherein, for said at least one template T j :

by each computation device:

receiving (S1) said biometric sample P of the user to be recognized,

for each template share

T

i

j

of said template stored by said computation device, computing (S2) a share of a global scoring

S

i

(

T

i

j

,

P

)

by computing a similarity between said stored template share and the received biometric sample P,

by said global scoring computation device, computing (S3) a global scoring for said at least one template by combining said shares of the global scoring computed for said template by said plurality of computation devices according to the formula:

(

T

j

,

P

)

=

i

=

0

n

-

1

S

i

(

T

i

j

,

P

)

mod

N

,

by said recognition device, recognizing (S4) said user to be recognized based on said global scorings computed for said at least one template.

2 . The method of claim 1 , wherein said biometric recognition comprises a recognition of a physiological characteristic of the user among a face, an iris, a fingerprint, a hand geometry, a voice pattern, a retinal pattern.

3 . The method of claim 1 , wherein the biometric sample of the user is among a picture of the user, a photo, a handwriting, speech, a fingerprint sample.

4 . The method of claim 1 wherein computing a similarity between said stored template share and the received biometric sample P comprises computing a cosine distance between the stored template share and the received biometric sample.

5 . The method of claim 1 , wherein the user is recognized by sorting global scorings computed for said at least one template and by selecting an identity corresponding to a template among said plurality of templates based on the sorted global scorings.

6 . The method of the previous claim 1 , comprising selecting an identity corresponding to the template having the highest global scoring.

7 . A computer program product directly loadable into the memory of at least one computer, comprising software code instructions for performing the steps below, when said product is run on the computer:

securing a biometric recognition of a user to be recognized against one or more templates T j , with j in [0, J−1] and J a non zero integer, corresponding to registered users identities;

using a biometric sample of said user, in a biometric recognition system comprising a plurality of computation devices, a global scoring computation device and a recognition device;

splitting each of said templates using an additive splitting modulo a predetermined integer N, into a plurality of template shares

T

i

j

stored in said computation devices, with i in [0, n−1] and n a non zero integer,

wherein, for said at least one template T j :

by each computation device:

receiving (S1) said biometric sample P of the user to be recognized,

for each template share

T

i

j

of said template stored by said computation device, computing (S2) a share of a global scoring

S

i

(

T

i

j

,

P

)

by computing a similarity between said stored template share and the received biometric sample P,

by said global scoring computation device, computing (S3) a global scoring for said at least one template by combining said shares of the global scoring computed for said template by said plurality of computation devices according to the formula:

(

T

j

,

P

)

=

i

=

0

n

-

1

S

i

(

T

i

j

,

P

)

mod

N

,

by said recognition device, recognizing (S4) said user to be recognized based on said global scorings computed for said at least one template.

8 . A biometric recognition system comprising

a plurality of computation devices;

a global scoring computation device; and

a recognition device ( 105 ), said system being configured for securing a biometric recognition of a user to be recognized against one or more templates T j , with j in [0, J−1] and J a non zero integer, corresponding to registered users identities, using a biometric sample of said user, wherein each of said templates is split, using an additive splitting modulo a predetermined integer N, into a plurality of template shares

T

i

J

stored in said computation devices, with i in [0, n−1] and n a non zero integer, and wherein, for said at least one template T j :

each computation device is configured for:

receiving said biometric sample P of the user to be recognized,

for each template share

T

i

j

of said template stored by said computation device, computing a share of a global scoring

S

i

(

T

i

j

,

P

)

by computing a similarity between said stored template share and the received biometric sample P,

said global scoring computation device is configured for computing a global scoring for said at least one template by combining said shares of the global scoring computed for said template by said plurality of computation devices according to the formula:

(

T

j

,

P

)

=

i

=

0

n

-

1

S

i

(

T

i

j

,

P

)

mod

N

,

said recognition device is configured for recognizing said user to be recognized based on said global scorings computed for said at least one template.

9 . The biometric recognition system of claim 8 , wherein the computation device is configured to be connected to a network and comprising a processor, a memory and an input-output interface.

10 . The biometric recognition system of claim 8 , wherein the global scoring computation device is configured to be connected to a network and comprising a processor, a memory and an input-output interface.

Assignments (2)
ASSIGNMENT OF ASSIGNOR'S INTEREST Recorded May 13, 2025
From: VACEK, JAN
To: THALES DIS CZECH REPUBLIC SRO.
Reel/Frame 071094/0958 →
ASSIGNMENT OF ASSIGNOR'S INTEREST Recorded May 13, 2025
From: THALES DIS CZECH REPUBLIC S.R.O.
To: THALES DIS FRANCE SAS
Reel/Frame 071095/0063 →
Priority Claims (1)
EP 22305557 · Apr 14, 2022 · regional
Continuity (1)
Related Publication 20250258898A1 · Aug 14, 2025
References Cited (27)
US 9935947B1 · Machani · 2018 [cited by examiner]
US 10452826B2 · Rush · 2019 [cited by examiner]
US 11223478B2 · Eldefrawy · 2022 [cited by examiner]
US 11991282B2 · Nagaraja · 2024 [cited by examiner]
US 11995884B1 · Sammoura · 2024 [cited by examiner]
US 12474231B2 · Taghavi · 2025 [cited by examiner]
US 20090161924A1 · Lu · 2009 [cited by examiner]
US 20120204035A1 · Camenisch · 2012 [cited by examiner]
US 20160148039A1 · Potash · 2016 [cited by examiner]
US 20170134375A1 · Wagner · 2017 [cited by examiner]
US 20170324736A1 · Connell, II · 2017 [cited by examiner]
US 20180129797A1 · Rush · 2018 [cited by examiner]
US 20180137382A1 · Nowak · 2018 [cited by examiner]
US 20180144176A1 · Ji · 2018 [cited by examiner]
US 20200184191A1 · Thompson · 2020 [cited by examiner]
US 20210014063A1 · Arora · 2021 [cited by examiner]
US 20210336792A1 · Agrawal · 2021 [cited by examiner]
US 20230017234A1 · Soryal · 2023 [cited by examiner]
US 20230050481A1 · Nagaraja · 2023 [cited by examiner]
US 20230084897A1 · Ryoo · 2023 [cited by examiner]
US 20240143717A1 · Baek · 2024 [cited by examiner]
US 20250258898A1 · Vacek · 2025 [cited by examiner]
US 20260038647A1 · Madani Tonekaboni · 2026 [cited by examiner]
WO 20210167534A1 · 2021 [cited by applicant]
International Search Report (PCT/ISA/220) and Written Opinion (PCT/ISA/237) mailed on Jun. 30, 2023, by the European Patent Office as the International Searching Authority for current International Application No. PCT/E… [cited by applicant]
XP055291441—Jun. 11, 2014—Julien Bringer et al Relevant to .Abstrat, Section 1.3, Section 2.3, 2.4, Section 3.3. [cited by applicant]
XP061041646—Nov. 4, 2020—Jia-Chng Loh et al Relevant to abstract. [cited by applicant]